Hugging Face's logo

thrunlab
/
original_glue_boolq

Transformers
Safetensors
mistral
trl
sft
Generated from Trainer
text-generation-inference
Inference Endpoints
Model card Files Community
lukeleeai commited on
Commit
c60b04e
•
1 Parent(s): c6b3bbf

End of training

Browse files
Files changed (4) hide show
  1. README.md +27 -28
  2. model-00001-of-00003.safetensors +1 -1
  3. model-00002-of-00003.safetensors +1 -1
  4. model-00003-of-00003.safetensors +1 -1
README.md CHANGED
@@ -21,8 +21,8 @@ should probably proofread and complete it, then remove this comment. -->
21
 
22
  This model is a fine-tuned version of [mistralai/Mistral-7B-Instruct-v0.1](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.1) on the super_glue dataset.
23
  It achieves the following results on the evaluation set:
24
- - Loss: 0.3093
25
- - Accuracy: 0.8707
26
 
27
  ## Model description
28
 
@@ -44,7 +44,7 @@ The following hyperparameters were used during training:
44
  - learning_rate: 1e-05
45
  - train_batch_size: 2
46
  - eval_batch_size: 4
47
- - seed: 0
48
  - distributed_type: multi-GPU
49
  - num_devices: 2
50
  - gradient_accumulation_steps: 2
@@ -58,31 +58,30 @@ The following hyperparameters were used during training:
58
 
59
  | Training Loss | Epoch | Step | Validation Loss | Accuracy |
60
  |:-------------:|:-----:|:----:|:---------------:|:--------:|
61
- | 0.5295 | 0.05 | 50 | 0.4429 | 0.8134 |
62
- | 0.3797 | 0.1 | 100 | 0.4171 | 0.8269 |
63
- | 0.1387 | 0.15 | 150 | 0.4193 | 0.8325 |
64
- | 0.3203 | 0.2 | 200 | 0.3964 | 0.8417 |
65
- | 0.2488 | 0.25 | 250 | 0.3915 | 0.8389 |
66
- | 0.2149 | 0.3 | 300 | 0.3729 | 0.8438 |
67
- | 0.2169 | 0.35 | 350 | 0.3606 | 0.8601 |
68
- | 0.4027 | 0.4 | 400 | 0.3680 | 0.8657 |
69
- | 0.3494 | 0.45 | 450 | 0.4100 | 0.8601 |
70
- | 0.1724 | 0.5 | 500 | 0.3634 | 0.8643 |
71
- | 0.346 | 0.55 | 550 | 0.4234 | 0.8671 |
72
- | 0.2071 | 0.6 | 600 | 0.3717 | 0.8629 |
73
- | 0.2983 | 0.65 | 650 | 0.3482 | 0.8622 |
74
- | 0.1517 | 0.7 | 700 | 0.3576 | 0.8728 |
75
- | 0.4414 | 0.75 | 750 | 0.3222 | 0.8650 |
76
- | 0.4264 | 0.8 | 800 | 0.3564 | 0.8820 |
77
- | 0.1045 | 0.85 | 850 | 0.3393 | 0.8721 |
78
- | 0.4414 | 0.9 | 900 | 0.3743 | 0.8671 |
79
- | 0.4848 | 0.95 | 950 | 0.3192 | 0.8657 |
80
- | 0.3788 | 1.0 | 1000 | 0.3339 | 0.8756 |
81
- | 0.6698 | 1.05 | 1050 | 0.3449 | 0.8728 |
82
- | 0.1779 | 1.1 | 1100 | 0.3368 | 0.8792 |
83
- | 0.397 | 1.15 | 1150 | 0.3678 | 0.8792 |
84
- | 0.0648 | 1.2 | 1200 | 0.3712 | 0.8749 |
85
- | 0.2106 | 1.25 | 1250 | 0.3511 | 0.8721 |
86
 
87
 
88
  ### Framework versions
 
21
 
22
  This model is a fine-tuned version of [mistralai/Mistral-7B-Instruct-v0.1](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.1) on the super_glue dataset.
23
  It achieves the following results on the evaluation set:
24
+ - Loss: 0.3115
25
+ - Accuracy: 0.8735
26
 
27
  ## Model description
28
 
 
44
  - learning_rate: 1e-05
45
  - train_batch_size: 2
46
  - eval_batch_size: 4
47
+ - seed: 1
48
  - distributed_type: multi-GPU
49
  - num_devices: 2
50
  - gradient_accumulation_steps: 2
 
58
 
59
  | Training Loss | Epoch | Step | Validation Loss | Accuracy |
60
  |:-------------:|:-----:|:----:|:---------------:|:--------:|
61
+ | 0.369 | 0.05 | 50 | 0.4328 | 0.8120 |
62
+ | 0.2708 | 0.1 | 100 | 0.4051 | 0.8283 |
63
+ | 0.6276 | 0.15 | 150 | 0.4020 | 0.8452 |
64
+ | 0.4395 | 0.2 | 200 | 0.3671 | 0.8452 |
65
+ | 0.3282 | 0.25 | 250 | 0.3746 | 0.8445 |
66
+ | 0.2967 | 0.3 | 300 | 0.3557 | 0.8523 |
67
+ | 0.2483 | 0.35 | 350 | 0.3862 | 0.8622 |
68
+ | 0.384 | 0.4 | 400 | 0.3765 | 0.8565 |
69
+ | 0.334 | 0.45 | 450 | 0.3628 | 0.8601 |
70
+ | 0.2671 | 0.5 | 500 | 0.3290 | 0.8664 |
71
+ | 0.2478 | 0.55 | 550 | 0.3421 | 0.8650 |
72
+ | 0.1814 | 0.6 | 600 | 0.3233 | 0.8693 |
73
+ | 0.3332 | 0.65 | 650 | 0.3451 | 0.8728 |
74
+ | 0.2063 | 0.7 | 700 | 0.3709 | 0.8678 |
75
+ | 0.2614 | 0.75 | 750 | 0.3530 | 0.8763 |
76
+ | 0.4273 | 0.8 | 800 | 0.3383 | 0.8721 |
77
+ | 0.1319 | 0.85 | 850 | 0.3360 | 0.8735 |
78
+ | 0.196 | 0.9 | 900 | 0.3096 | 0.8735 |
79
+ | 0.3564 | 0.95 | 950 | 0.3354 | 0.8770 |
80
+ | 0.3145 | 1.0 | 1000 | 0.3421 | 0.8784 |
81
+ | 0.1344 | 1.05 | 1050 | 0.4273 | 0.8735 |
82
+ | 0.4227 | 1.1 | 1100 | 0.3555 | 0.8707 |
83
+ | 0.1696 | 1.15 | 1150 | 0.3399 | 0.8742 |
84
+ | 0.5423 | 1.2 | 1200 | 0.3405 | 0.8813 |
 
85
 
86
 
87
  ### Framework versions
model-00001-of-00003.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:77bd6c46d43a529cb0b02f18114a54f915db6dc45007707cd70118305bf38f3c
3
  size 4943163992
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e019a5c835a08dc82fc4526058a1a1f832185a6aa75814f44319b088de4dcdab
3
  size 4943163992
model-00002-of-00003.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:f96b5186ae311378d272f07dc7cd6306cc4b5d971dcad6f8d29df5a9c5502816
3
  size 4999821144
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2ec161cb1d01a097e6c7c6655e1323d9fe2ac8f828151c9fce8cec20189dd520
3
  size 4999821144
model-00003-of-00003.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:54ae1cabb63f4fc65fb566649fda211fe62c64bdfe410c7ece8035dfc4cd0d04
3
  size 4540517840
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6e06a98b0bb7e099fc353489ada758a39a18535ec8f1278a2cddbc763d77e039
3
  size 4540517840